International Journal of Molecular Sciences Article Predominant Distribution of the RNAi Machinery at Apical Adherens Junctions in Colonic Epithelia Is Disrupted in Cancer Joyce Nair-Menon 1, Amanda C. Daulagala 1 , Dean M. Connor 2, Lauren Rutledge 1, Trevor Penix 1, Mary Catherine Bridges 1, Bridgette Wellslager 1, Demetri D. Spyropoulos 3 , Cynthia D. Timmers 4, Ann-Marie Broome 2 and Antonis Kourtidis 1,* 1 Department of Regenerative Medicine and Cell Biology, Medical University of South Carolina, 173 Ashley Avenue, Charleston, SC 29425, USA; [email protected] (J.N.-M.); [email protected] (A.C.D.); [email protected] (L.R.); [email protected] (T.P.); [email protected] (M.C.B.); [email protected] (B.W.) 2 Department of Cell and Molecular Pharmacology and Experimental Therapeutics, Medical University of South Carolina, 173 Ashley Avenue, Charleston, SC 29425, USA; [email protected] (D.M.C.); [email protected] (A.-M.B.) 3 Department of Pathology and Laboratory Medicine, Medical University of South Carolina, 173 Ashley Avenue, Charleston, SC 29425, USA; [email protected] 4 Department of Medicine, Medical University of South Carolina, 171 Ashley Avenue, Charleston, SC 29425, USA; [email protected] * Correspondence: [email protected]; Tel.: +1-843-792-9170 Received: 10 March 2020; Accepted: 4 April 2020; Published: 7 April 2020 Abstract: The RNA interference (RNAi) machinery is an essential component of the cell, regulating miRNA biogenesis and function. RNAi complexes were thought to localize either in the nucleus, such as the microprocessor, or in the cytoplasm, such as the RNA-induced silencing complex (RISC). We recently revealed that the core microprocessor components DROSHA and DGCR8, as well as the main components of RISC, including Ago2, also associate with the apical adherens junctions of well-differentiated cultured epithelial cells. Here, we demonstrate that the localization of the core RNAi components is specific and predominant at apical areas of cell-cell contact of human normal colon epithelial tissues and normal primary colon epithelial cells. Importantly, the apical junctional localization of RNAi proteins is disrupted or lost in human colon tumors and in poorly differentiated colon cancer cell lines, correlating with the dysregulation of the adherens junction component PLEKHA7. We show that the restoration of PLEKHA7 expression at adherens junctions of aggressively tumorigenic colon cancer cells restores the junctional localization of RNAi components and suppresses cancer cell growth in vitro and in vivo. In summary, this work identifies the apical junctional localization of the RNAi machinery as a key feature of the differentiated colonic epithelium, with a putative tumor suppressing function. Keywords: E-cadherin; p120 catenin; PLEKHA7; DROSHA; DGCR8; Ago2; GW182; RNA interference; colorectal; tumor suppressor 1. Introduction Colorectal cancer is the third most prevalent and second deadliest form of cancer [1]. This implies that there are still gaps in our understanding of the disease [2]. A common feature of colorectal tumors is disruption and eventual loss of epithelial tissue architecture and integrity, which occurs in early or Int. J. Mol. Sci. 2020, 21, 2559; doi:10.3390/ijms21072559 www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2020, 21, 2559 2 of 20 even pre-cancerous stages of the disease [3,4]. However, the extent that disruption of colonic epithelial architecture contributes to pro-tumorigenic cell transformation is still unclear [2–5]. The adherens junction is a cell-cell adhesion structure critical for organ development, tissue architecture, and cell function [6,7]. E-cadherin is the core adhesion molecule in epithelial cells and is stabilized through its association with p120 catenin (p120) [8]. Due to its essential function in epithelial homeostasis, E-cadherin has long been considered a tumor suppressor, which is downregulated during epithelial-mesenchymal transition, promoting tumor progression and metastasis. However, numerous studies in recent years have challenged this notion [5,9]. For example, we and others have shown that E-cadherin is still expressed in most tumors examined and that E-cadherin-p120 complexes are required for cancer cell survival, proliferation, anchorage-independent growth, as well as for collective cell migration [5,10–18]. These findings complicated our understanding of the function of E-cadherin complexes in tumorigenesis. Although E-cadherin-based cell-cell contacts form across lateral membranes between adjacent cells, mature adherens junctions appear in well-differentiated epithelia and specifically localize at the apical areas of cell-cell contact, where they tether to circumferential actin to form the zonula adherens [6,7]. This and other findings point toward structural differences between the apical adherens junctions and basolateral areas of cell-cell contact. For example, it has been shown that a p120-binding partner called PLEKHA7 specifically localizes at the apical adherens junctions of well-differentiated epithelial cells and tissues, but not at the lateral areas of cell-cell contact [10,19,20]. In addition to these structural differences between the apical and basolateral complexes at areas of cell-cell contact, we recently showed that these complexes are also functionally distinct. In particular, we demonstrated that the apical adherens junctions can suppress pro-tumorigenic cell transformation through PLEKHA7, whereas the basolateral cell-cell junction complexes promote it, in the absence of PLEKHA7 [10,21,22]. In agreement with these findings, we and others have shown that PLEKHA7 is broadly mis-localized or downregulated in human breast and kidney tumors, although E-cadherin is still broadly expressed at areas of cell-cell contact [10,23]. These findings provided a mechanistic basis to explain why E-cadherin complexes can act both in a pro- and in an anti-tumorigenic fashion and revealed that other cadherin-junction components have to be taken into consideration, when assessing the role of E-cadherin in tumorigenesis. Our investigation to understand how the adherens junctions may be acting as a tumor suppressor led to the revelation that PLEKHA7 recruits numerous RNA-binding proteins, including the main complexes of the RNA interference (RNAi) machinery [10,24,25]. More specifically, we showed that PLEKHA7 recruits the core components of the microprocessor complex, namely DROSHA and DGCR8, as well as the RNA-induced silencing complex (RISC), including its key enzymatic component Ago2 [10,25]. Notably, we showed that this interaction occurs specifically at the apical adherens junctions of well-differentiated epithelial cells, such as human colon Caco2 and canine kidney MDCK cells [10,25]. By recruiting these complexes, PLEKHA7 regulates the processing and activity of a number of tumor-suppressing miRNAs, such as miR-30b, miR-24, miR-200c, and miR-203a [10,25]. PLEKHA7 depletion in Caco2 cells results in the disruption of the localization of the RNAi machinery at the adherens junctions and in decreased levels and activity of these miRNAs, eventually promoting oncogene upregulation and pro-tumorigenic transformation of Caco2 cells [10,25]. These findings provided a path to explain the function of the apical adherens junction complex as a suppressor of pro-tumorigenic cell transformation. In addition, these results challenged the broadly accepted notion that the microprocessor complex localizes solely in the cell nucleus and that the RISC specifically localizes in the cytoplasm, by demonstrating their additional association with the apical adherens junctions of well-differentiated epithelial cells [24,26,27]. Since we obtained these data in cultured epithelial cells, such as colon Caco2 cells, the question that emerged is whether the junctional localization of the RNAi machinery is broad and physiologically relevant to other cells and tissues, either normal or diseased. In this work, we seek to examine whether this subcellular distribution of the microprocessor and RISC at apical adherens junctions is indeed a feature of the well-differentiated colonic epithelium. Int. J. Mol. Sci. 2020, 21, 2559 3 of 20 We also examine whether this localization is maintained or lost in colon cancer cells and tumors, since our findings point towards a putative tumor suppressing mechanism that is potentially disrupted in cancer.Int. In J. addition, Mol. Sci. 2020 we, 21,test x FOR whether PEER REVIEW we can restore this junctional localization of RNAi in colon3 of 20 cancer cell lines and look at effects on their growth. Our findings presented here confirmed our predictions, mechanism that is potentially disrupted in cancer. In addition, we test whether we can restore this indicating that the apical adherens junction-associated RNAi machinery is a predominant feature of junctional localization of RNAi in colon cancer cell lines and look at effects on their growth. Our well-difffindingserentiated presented colonic here epithelia, confirmed which our predictions, is broadly disruptedindicating that incolon the apical cancer. adherens junction- associated RNAi machinery is a predominant feature of well-differentiated colonic epithelia, which 2. Resultsis broadly disrupted in colon cancer. 2.1. The2. Core Results Components of the RNAi Machinery Localize at Apical Adherens Junctions in Normal Colon Epithelia We2.1. have The Core recently Components shown of thatthe RNAi the Machinery adherens Loca junctionslize at Apical recruit Adherens the coreJunctions components in Normal Colon of